Agentic Security Engineer (LLM agent red teaming)
You test LLM agents against attack, then help limit what they can reach. You plan and run red-team exercises on agents, their tools and their data access, on systems the client owns or is authorized to test and within a written scope. Every finding you report comes with the run that shows it.
Duties
- Agree the scope of each engagement in writing: the agents, the tools and data they reach, and where they run.
- Run direct and indirect prompt injection attacks through pages, emails, files and tool results.
- Test whether an agent can be steered into calling a tool it should not call, or with arguments it should not send.
- Review MCP server permissions, tool descriptions and credentials for access beyond what each task needs.
- Trace the paths by which private data in an agent's context or tools could leave through its answers or tool calls.
- Test whether content planted in memory or retrieved documents, or pressure from another agent, changes what an agent does.
- Run every attack on pinned model versions and log each run so it can be replayed.
- Work with the client's team to close each finding, from narrower credentials to new approval steps.
- Rerun every attack that succeeded to confirm each fix, and write findings a developer can reproduce.
What you bring
- 3+ years in application security, red teaming or security engineering.
- Hands-on experience attacking LLM applications or agents, prompt injection included.
- Working knowledge of how agents call tools, including MCP and function calling.
- Python, and comfort reading the code of the systems you test.
- Clear written findings, each with the steps that reproduce it.
Nice to have
- Experience with OAuth, scoped credentials and secrets handling.
- Experience with Kubernetes access control and network policies.
- Familiarity with the OWASP Top 10 for LLM Applications.
- Published research or talks on attacks against LLM agents.
Pay and benefits
$123,000 to $190,000 a year
Full-time W-2 employment with health insurance, a retirement plan and paid leave.
No application deadline. We accept applications for this role on an ongoing basis.
Location
Remote, United States
Apply
We review applications as each team grows and contact you when there is a match.
Rillor is an equal opportunity employer. We consider qualified applicants without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, age, disability, veteran status or any other status protected by law.
If you need an accommodation to apply, email careers@rillor.com.
Notices
Trademarks. NVIDIA, AMD, Ceph, Kubernetes and related product names are trademarks of their owners and are used only to identify products. No affiliation or endorsement is implied.